针对高级量测体系中的海量数据问题,首次引入压缩感知以克服传统数据压缩方法的不足,深入探索了基于压缩感知的高级量测体系(advanced metering infrastructure based on compressed sensing,AMI-CS).首先,在分析各类数据特点的基础上,提出了基于时间和基于空间的 2 种基本模型及其选取原则;然后,设计模型中的关键要素,提出分类K-SVD稀疏基和适用于时间模型的优选重构算法,并设置二进稀疏测量方式、通用重构算法及适用采集参数;基于此,形成了AMI-CS具体构建方案.实验结果表明,所提出的AMI-CS方案关键要素均具合理性,优于CS传统要素且较传统压缩提升了抗丢包性,通过合理选择压缩比,数据重构信噪比在 58 dB以上、重构误差在0.24%以下,满足AMI要求.
Advanced Metering Infrastructure Based on Compressed Sensing in Smart Grid
To solve massive data issues for advanced metering infrastructure(AMI),this paper introduces a compressed sensing(CS)technique to overcome the deficiencies of traditional data compression method and explore the advanced metering infrastructure based on compressed sensing(AMI-CS).Firstly,on the basis of analyzing the characteristics of various types of AMI data,this paper proposes two basic models based on time and based on space as well as their selec-tion principles.Then,the key elements of the construction scheme are designed,a classified K-SVD sparse basis and a optimal reconstruction algorithm are proposed.Binary sparse matrix is set as measurement way while reconstructing al-gorithm and acquisition parameters are set.A specific construction scheme for AMI-CS is formulated and the experimental results show that key elements in scheme are all reasonable.By choosing compression ratio reasonably,the data reconstruction signal-to-noise ratio is above 58 dB and the reconstruction error is below 0.24%,which meet require-ments of AMI.
国家科技期刊平台
NSTL
万方数据
